14th International Myeloma Workshop

Whole genome sequencing reveals underlying genetics of multiple myeloma

Dr Jonathan Keats – The Translational Genomics Research Institute (TGen), Phoenix, USA

The talk I gave was about a lot of the whole genome sequencing initiatives that have occurred in multiple myeloma that I’ve been directly a part of, though that represents a large majority of what’s going on in multiple myeloma today.

What have you been looking at?

We have had a couple of different studies, mostly whole exome sequencing in multiple myeloma, and now we’re moving into situations where we’ve paired whole exome sequencing with RNA sequencing and whole genome sequencing for what we would say is the first time in a given myeloma patient we can truly characterise every genetic event that has occurred in that patient. So it really is the first time we’ve ever been able to completely understand the genetic basis for that patient’s cancer and we now can now build on that, hopefully, in the future.

How many patients have you looked at?

From that comprehensive study now we’ve completed almost 260 in total to date and that’s increasing now over the next three years to be over 1200.

At a number of different institutes?

Yes, I guess we’re verging on 61 different clinical centres that are submitting samples. We have one biorepository that supports the entire operation and then one centre that does all the sequencing, our centre at the Translational Genomics Research Institute.

What have you found?

The most common things that are things that we know are important in cancer in general, have been KRAS and NRAS which are very common in all cancers. But in myeloma what has been the most interesting and valuable from the original parts of these studies has been identifying two genes, one called FAM46C and the other one DIS3 that are clearly very important in the disease but we really don’t understand at all what they contribute.  It’s confusing because somebody always asks, “Well, what do they do?” and we say we don’t know. It’s also good because a lot of other studies, say in breast cancer and brain cancer, did similar things but didn’t find anything new of any common high frequency. So at least in the field of myeloma we do have some things that have the potential for new therapeutics, things like Ras we’ve spent thirty years trying to develop therapeutics on and hopefully in the next thirty years we’ll find one but at least having something that looks like it’s very important in our disease will give us the opportunity to potentially have therapeutics for those; maybe it will be easier or more druggable.

That’s a long process?

The process for those things are shortening as technology improves, drug companies have bigger resources and we improve screening methodologies.

And a number of subgroups have been identified?

That has largely been something that we haven’t dramatically improved over what we had from microarray days, that switching to a sequencing based test, as opposed to a microarray-based test, has generally recapitulated the same groups; we haven’t really identified dramatically new groups. The advantage has been that we can now integrate different data sources to break those subgroups down into additional subgroups that have more clinical utility is what it looks like to date.

Where does this research hopefully get us?

The hope, especially at our institute, that we’re moving to is the ability to do whole genome sequencing on each individual patient and, based on those results, actually predict therapies. So today an average myeloma patient really gets a regimen that usually involves either an immunomodulatory drug or a proteasome inhibitor or maybe both and then potentially going on to a transplant and the hope is to be able to move into a time point where you go, “You know, for this patient we should be adding a BRAF inhibitor, because that patient actually does have a BRAF mutation, to our standard regimen. And/or maybe in this case the patient has a p10 abnormality suggesting that we should add something that hits the PI3 kinase pathway.” So we’ll start personalising, or what we call precision medicine, really going after what is wrong with that tumour, not necessarily treating you as a multiple myeloma patient but treating your tumour specifically.

But this is more work and cost?

It is because of the additional screening that will be involved though those costs are coming down dramatically. In the research setting, like in a large institute like ours, we are moving to doing this on a regular basis for a lot of our large studies, our clinical trials, where we are adding that type of information to the patient’s therapeutic options. But again those are not necessarily where things are today in the average clinical patient but part of that is really limited to the proof that there’s value to doing that, which we believe we will prove. Then, more importantly, the dissemination of the technology and ease of use of analysis, particularly of the sequencing based data, still does require significant infrastructure but that is changing rapidly as the costs of sequencing and methodology improves. Our expectation is probably within five years, especially in academic centres, this will probably be standard of care within the next five to ten years for sure.

What has your institute brought to this field?

The biggest thing that we’ve added in the myeloma field, I think, makes us a bit unique compared to others is we’ve taken all the available model systems and done equal characterisation. The hope there is to accelerate these ideas of how do we take what we’ve found in patients to identify therapeutics is now really having a better understanding of the available models, we can really model which drugs or test drugs that meet certain genetic backgrounds. So if we know you have NF-kappaB activation plus Ras pathway activation, what are the drugs that work in that background versus if you have just Ras without NF-kappaB, that may lead to different combinations and we’re going to be lucky compared to all the other fields because we do have this resource of having all of our models equally characterised as our patients. So I think that will hopefully accelerate how fast we can get these things back to patients.